COUGAR: The Network Is The Database
The widespread distribution and availability of small-scale sensors, actuators,
and embedded processors is transforming the physical world into a computing
platform. Sensor networks that combine physical sensing capabilities such as
temperature, light, or seismic sensors with networking and computation
capabilities will soon become ubiquitous. Applications range from environmental
control, warehouse inventory, and health care to scientific and military
Existing sensor networks assume that the sensors are preprogrammed and send data to a central frontend where the data is aggregated and stored for offline querying and analysis. This approach has two major drawbacks. First, the user cannot change the behavior of the system dynamically. Second, communication in today's networks is orders of magnitude more expensive than local computation; thus in-network storage and processing can vastly reduce resource usage and extend the lifetime of a sensor network.
Sensor nodes come in a variety of hardware configurations, from nodes connected to the local LAN attached to permanent power sources to nodes communicating via wireless multi-hop RF radio powered by small batteries, the types of sensor nodes following resource constraints:
- Communication: The wireless network connecting the sensor nodes provides usually only a very limited quality of service, has latency with high variance, limited bandwidth, and frequently drops packets.
- Power consumption: Sensor nodes have limited supply of energy, and thus energy conservation needs to be of the main system design considerations of any sensor network application.
- Computation: Sensor nodes have limited computing power and memory sizes. This restricts the types of data processing algorithms on a sensor node, and it restricts the sizes of intermediate results that can be stored on the sensor nodes.
- Uncertainty in sensor readings: Signals detected at physical sensors have inherent uncertainty, and they may contain noise from the environment. Sensor malfunction might generate inaccurate data, and unfortunate sensor placement (such as a temperature sensor directly next to the air conditioner) might bias individual readings.
We investigate two unique approaches to sensor networks. First, we will use a
database approach to unite the seemingly conflicting requirements of
scalability and flexibility in monitoring the physical world. The objective of
this research is to build a new distributed data management layer that scales
with the growth of sensor interconnectivity and computational power on the
sensors over the next decades. Our system will reside directly on the sensor
nodes and create the abstraction of a single processing node without
centralizing data or computation. The system will provide scalable,
fault-tolerant, flexible data access and intelligent data reduction, and its
design involves a confluence of novel research in database query processing,
networking, algorithms, and distributed systems.
Second, we believe that due to the heavily resource-constrained environment of sensor networks, cross-layer optimizations allow interesting opportunities for the preservation of resources. Due to the regularity of query processing patterns we believe that we can design query-layer specific routing algorithms that are optimized --- not for general point-to-point communication --- but for the more regular types of communication patterns that are generated by a query layer. Investigation of such cross-layer optimizations is the second major goal of this research.
From a research standpoint, the central issues are the following:
- Sensor networks as a distributed database system. What is the impact of storing data in the networks for later querying? How do we optimize and process declarative queries involving sensor data? Sensors have limited battery power and the wireless network has limited bandwidth and quality of service; query execution has to take these constraints into account.
- Cross-layer optimizations. How can we expose structure that originates in the data management layer to the routing layer? Should we design a data management layer that is optimized for a given routing layer, or should we optimize the routing layer given the data management layer? What are suitable interfaces that enable this tight coupling?
BOOM is a showcase for student research at Cornell. See the BOOM 2003 website for more information.
Jay Ayres (Junior, Computer Science)
Nick Gerner (Sophomore, Computer Science)
Joel Ossher (Freshman, Computer Science)
Lin Zhu (Senior, Computer Science)
Sensoria Node Demo
The Sensoria Demo shows how Sensoria nodes can be used to track moving objects.
A Tracking GUI is used to specify simulated motion tracks, and then the main COUGAR GUI detects these motion tracks and display them onscreen. In addition, we will describe how the Sensoria nodes can be used in conjunction with Mica motes to form a large sensor network.
The demonstration slides are available here .
Mica Mote Demo
This demo shows how a network of Mica motes can be used to retrieve
environmental data such as light and temperature readings.
One important feature of our network is its use of geographical routing, where each mote knows its neighbors and is able to efficiently direct messages to those neighbors. The network includes a protocol for the motes to continuously poll its neighbors to see if a link in the network has been broken.
Another feature is the usage of in-memory databases on each of the motes. Tables in these databases can be used to store a mote's neighbors, recent sensor readings that can be aggregated, or a mote's location. The database system is extremely flexible, allowing for future functionality additions.
The demonstration slides are now online here.
- Graduate students: Yong Yao
- Faculty: Al Demers, Johannes Gehrke, Rajmohan Rajaraman (Northeastern University), Sergio Servetto
Alumni and their first location after Cornell
- Jay Ayres (graduate school at Stanford)
- Philippe Bonnet (faculty at the University of Copenhagen)
- Zhiyuan Chen (Microsoft)
- Wai Fu Fung
- Nick Gerner (Microsoft)
- Tobias Mayr (IBM Almaden Research Center)
- David Sun
- Ben Szekely (IBM)
- Joel Ossher (graduate school at UMass)
- Lin Zhu
- Niki Trigoni , Yong Yao , Alan J. Demers , Johannes Gehrke, Rajmohan Rajaraman : Multi-query Optimization for Sensor Networks. DCOSS 2005 : 307-321
- Alan J. Demers , Johannes Gehrke, Mingsheng Hong , Mirek Riedewald : Processing High-Speed Intelligence Feeds in Real-Time. ISI 2005 : 617-618
- Prakash Linga , Adina Crainiceanu , Johannes Gehrke, Jayavel Shanmugasundaram : Guaranteeing Correctness and Availability in P2P Range Indices. SIGMOD Conference 2005 : 323-334
- Niki Trigoni , Yong Yao , Alan J. Demers , Johannes Gehrke, Rajmohan Rajaraman : Hybrid Push-Pull Query Processing for Sensor Networks. GI Jahrestagung (2) 2004 : 370-374
- Niki Trigoni, Yong Yao, Alan Demers, Johannes Gehrke, Rajmohan Rajaraman . WaveScheduling: Energy-Efficient Data Dissemination for Sensor Networks . In International Workshop on Data Management for Sensor Networks (DMSN), in conjunction with VLDB, 2004.
- Samuel Madden and Johannes Gehrke. Query Processing in Sensor Networks . IEEE Pervasive Computing, Vol. 3, No. 1., January-March March 2004.
- David Kempe, Alin Dobra, and J. E. Gehrke. Computing Aggregate Information using Gossip . In Proceedings of the 44th Annual IEEE Symposium on Foundations of Computer Science (FOCS 2003) . Cambridge, MA, October 2003.
- Yong Yao, J. E. Gehrke . Query Processing in Sensor Networks. In Proceedings of the First Biennial Conference on Innovative Data Systems Research (CIDR 2003). Asilomar, California, January 2003.
- Yong Yao, J. E. Gehrke . The Cougar Approach to In-Network Query Processing in Sensor Networks. Sigmod Record, Volume 31, Number 3. September 2002.
- Anton Faradjian, J. E. Gehrke, and Philippe Bonnet. GADT: A Probability Space ADT For Representing and Querying the Physical World. Proceedings of the 18th International Conference on Data Engineering . San Jose, California, February 2002.
- Philippe Bonnet, J. E. Gehrke, and Praveen Seshadri . Towards Sensor Database Systems . Proceedings of the Second International Conference on Mobile Data Management . Hong Kong, January 2001.
- Philippe Bonnet, J. E. Gehrke, and Praveen Seshadri . Querying the Physical World . IEEE Personal Communications, Vol. 7, No. 5, October 2000, pages 10-15. Special Issue on Smart Spaces and Environments.
- Philippe Bonnet, Praveen Seshadri . Device Database Systems. In Proceedings of the 16th International Conference on Data Engineering . San Diego, California, February 2000.
The COUGAR Project is currently supported by two grants from the National Science Foundation: CAREER Grant IIS-0133481 and Grant IIS-0330201 (joint with Northeastern University). Any opinions, findings and conclusions or recomendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation (NSF).
We also collaborate with the Cornell Information Assurance Institute .
The early work on Cougar has been supported by the Defense Advanced Research Project Agency and by a gift from Intel. Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of these sponsors.